Why Do Great Relievers Seem So Bad?
The world's best relievers are losing lots of games this month.
The dream for any team is to have a closer who pitches like Mariano Rivera in the postseason. The best way to make this dream come true is to acquire (or develop) a closer who pitches like Mariano Rivera in the regular season.
The problem, as team after team and fan after fan have found, is that lots of pitchers (relatively speaking) can pitch like Rivera in the regular season, while it’s as inevitable as death that none pitches like Rivera in the postseason. In just the past week, we’ve seen all four of this era’s Big Four closers fail fantastically:
On Monday, the Guardians’ Emmanuel Clase allowed three straight hits in a tie game, including a game-losing home run—the hardest-hit ball he’d ever allowed.
On Sunday, the Mets’ Edwin Díaz allowed three runs in two-thirds of an inning, blowing a one-run lead.
On Thursday, the Brewers’ Devin Williams blew a 2-0 lead in the ninth, allowing four runs in two-thirds of an inning.
On Wednesday, the Astros’ Josh Hader allowed five of the eight batters he faced to reach, and the 2-2 tie he inherited became a 5-2 Astros loss.1
It’s not crazy to call those four pitchers Riveraesque from April through September. Since 2018, they have won eight of the 12 Reliever of the Year awards, and Clase will certainly win it this season. Collectively,
Those four dudes’ careers, averaged: 195 ERA+, 2.58 FIP
Mariano Rivera, career: 205 ERA+, 2.76 FIP
Yet it doesn’t seem to matter how good pitchers are in the regular season. Come the fall,
they ain’t Mo.
*****
Consider:
In a world without Rivera, Billy Wagner might be remembered as the greatest regular-season reliever of all-time. His career postseason ERA was 10.03 (in 14 appearances).
Before his very-late-career numbers gutted him, Craig Kimbrel had been even better than Wagner in the regular season. But his career postseason ERA is 4.50 in 30 appearances. His postseason FIP, at 5.16, is exactly double his regular-season FIP, 2.58.
Hall of Famer Lee Smith allowed runs in three-quarters of his postseason outings. Hall of Famer Trevor Hoffman blew a third of his postseason saves—and every time he entered a tie game in the postseason (n=3) he left with his team behind.
What’s the first image your brain pulls up for Dennis Eckersley? It’s definitely Kirk Gibson yarding off him in the World Series. Of Goose Gossage? You might have something else, but for me it’s Kirk Gibson yarding off him in the World Series. Aroldis Chapman? Rajai Davis, yardo.
So there is a reason this tweet did pretty big numbers this week:
It feels right.
*****
Why are elite relievers so shaky in the postseason? Well, surprise: They’re actually not.
I know all the stuff I just put in front of you. But we can consider it a different way. Let’s filter for great reliever seasons. Let’s try something like this:
50+ innings, ERA below 2, FIP below 3, team made postseason
From 2004 through 2023, a 20-year period of modern reliever usage, there were 68 non-Rivera seasons that met those standards. There are a small handful of names in there that you’d flag as nonconvincingly upper tier—Dennys Reyes one year, Dan Otero one year—but that’s fine. That’s the point of relievers. In 60-inning increments, flukes slip in.
Even with the flukes, though, this group did extremely well in the postseasons that coincided with their elite seasons. Collectively, they threw 364 postseason innings, and they allowed 100 runs—not earned runs, but runs. That’s an average of 2.47 runs per nine innings. Mariano Rivera, in his regular-season career, allowed 2.38 runs per nine innings. The elite relievers were, in fact, still elite relievers.
So why do so many of us think the great relievers fail so often? There are, I think, five factors:
1. Availability bias.
The availability heuristic is “a type of cognitive bias that … involves relying on information that comes to mind quickly or is most available to us.”
A reliever’s greatest October failures will almost certainly come to mind more quickly than his greatest October successes. The failures are more unexpected (because elite closers usually get hitters out), they’re more dramatic (because changing the state of a drama is more dramatic than preserving the status quo), and they’re more likely to get repeated on highlights. You’ve seen Kirk Gibson chugging around the bases thousands of times; you’ve seen Eckersley getting the final three outs of the World Series 12 months later probably once.
So it’s easy to call to mind a pitcher’s worst moments. I have absolutely no doubt that Aroldis Chapman is shaky in the postseason, because I remember him wobbling in Game 7 of the 2016 World Series, and scowling after allowing Rajai Davis’ home run. In fact, Chapman has been fairly dominant in the postseason in his career, with a 2.37 ERA in 44 appearances. Since the Rajai Davis home run, that ERA is 1.86, and I’d struggle to even picture him in the right postseason uniforms in many of those years.
2. Negativity bias
Brad Lidge had a 2.18 ERA in 39 career postseason starts, and yet every October his name resurfaces as a punchline and a cautionary tale:
There’s a baseball axiom, or maybe two separate axioms that I’m mashing together, that says every fan thinks their team’s shortstop is Gold Glove caliber and every fan thinks their team’s closer is unreliable. The exception to the latter is Rivera, who only had one detractor (Ben Lindbergh’s mom). But almost every other closer ends up disproportionately defined by how you feel when you’re watching him (nervous, the game is on the line!) and by the salience of his failures.
Keep reading with a 7-day free trial
Subscribe to Pebble Hunting to keep reading this post and get 7 days of free access to the full post archives.